INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Vegetables contaminated with Listeria monocytogenes were responsible for the first verified food-mediated outbreak of listeriosis. More recently, cheese and meat products have been the sources of several outbreaks and sporadic cases (22). Seafoods have not been linked to large outbreaks of listeriosis, but ready-to-eat products of rainbow trout from one producer were found to be the source of a small outbreak in Sweden. Recently, five cases of febrile gastroenteritis in Finland were associated with cold smoked salmon contaminated with L. monocytogenes (7, 16). Other contaminated seafoods, such as smoked cod roe, mussels, and undercooked fish, have been assumed to be sources for sporadic cases of listeriosis (5, 8, 9).

L. monocytogenes is sporadically isolated from a range of seafoods, including ready-to-eat products like cold and hot smoked salmon, gravad salmon, shrimp, fermented fish, and fish salads (2). While the consumption of ready-to-eat seafoods is high in some countries, the low levels (<100 CFU/g) of L. monocytogenes normally detected may explain why only a few cases and minor outbreaks have been linked to these products. However, levels of more than 1,000 CFU/g have been found from time to time, and in screenings, up to 132,000 and 24,500 CFU/g have been isolated from hot smoked and cold smoked salmon, respectively (6, 12, 15, 21). The cold smoked rainbow trout associated with the Finnish outbreak contained 190,000 CFU/g (16).

To evaluate the role of seafoods in the epidemiology of listeriosis and to identify possible sources of the bacteria in the products, Boerlin et al. (4) used molecular typing methods to study L. monocytogenes isolated from fish products imported into Switzerland and compared these bacteria to those recovered from human patients. They reported that some clones were widespread in fish products but concluded that isolates of human origin and those from fish products belonged to separate subpopulations of the species.

In an investigation of the routes for L. monocytogenes contamination of the products in a smoked-salmon-processing plant in Norway, we have shown that one clone of L. monocytogenes colonized the processing environment during an 8-month sampling period and contaminated the smoked salmon during processing. This clone was the one most frequently recovered from human patients in Norway. The origins of the bacteria in the smokehouse were not identified, but raw fish was not assumed to be the source, as it was infrequently contaminated and only then with clones which were found neither in the processing environment nor in the final product (20).

The aim of the present study was to investigate if seafoods and fish-processing environments in general, or in specific plants, offer an ecological niche to certain L. monocytogenes clones and whether cases of listeriosis could be linked to contaminated seafoods. For this purpose, isolates from fish products and processing environments from various producers in Norway, as well as from patients, were characterized using multilocus enzyme electrophoresis (MEE) and restriction enzyme analysis (REA) of total DNA.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial isolates. A total of 305 L. monocytogenes isolates were examined. Forty-four isolates were recovered from human listeriosis patients in Norway between 1992 and 1996, and they included all L. monocytogenes human isolates forwarded to the National Institute of Public Health, Oslo, Norway, in that period. The remaining 261 isolates were collected between 1992 and 1997 from seafood products or the seafood-processing environment. Of these, 93 were obtained from a variety of seafood producers in Norway, with from one to four isolates from 55 individual producers or environmental sites. These isolates comprised (i) 36 isolates from ready-to-eat seafoods: cold smoked salmon (17 isolates), cold smoked mackerel (1 isolate), gravad salmon (5 isolates), fermented trout (3 isolates), marinated fish (6 isolates), shrimp (2 isolates), and brine for anchovies (2 isolates); (ii) nine isolates from raw seafoods: slaughtered fish (four isolates) or filleted fish (four isolates) or other raw seafoods (one isolate); and (iii) 48 environmental isolates, including isolates from the cold-smoking processing environment (31 isolates), salmon slaughterhouses (9 isolates), and filleting environments (2 isolates), as well as seawater (5 isolates) and seagull feces (1 isolate).

MEE analysis. Methods of protein extract preparation, starch gel electrophoresis, and selective enzyme staining were as described earlier (13, 23). Twenty-one enzymes were assayed: indophenol oxidase, 6-phosphogluconate dehydrogenase, glucose-6-phosphate dehydrogenase, adenylate kinase, phosphoglucose isomerase, two glutamate dehydrogenases, three peptidases, nucleoside phosphorylase, acid phosphatase, phosphoglucomutase, glyceraldehyde-3-phosphate dehydrogenase, isocitrate dehydrogenase, esterase, mannose phosphate isomerase, fumarase, alanine dehydrogenase, lactate dehydrogenase, and catalase. Multilocus enzyme genotypes were equated with clones. Genetic distance between pairs of electrophoretic types (ETs) was expressed as the proportion of enzyme loci at which dissimilar alleles occurred (mismatches), and clustering was performed from a matrix of genetic distances by the average-linkage method (24).

REA. To improve the discrimination, isolates of the same ET were further typed by REA using HaeIII. The bacteria were grown overnight at 37°C on blood agar plates. Preparation of genomic DNA, DNA digestion, and gel electrophoresis were performed as described earlier (17). Electrophoresis was accomplished using recirculation of 1× TAE buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.0). A mixture of HindIII-digested lambda DNA (New England BioLabs, Beverly, Mass.) and a 1-kb DNA ladder (Gibco BRL, Gaithersburg, Md.) was employed as the size marker. L. monocytogenes strain L 1349 was used as the control. Isolates with identical banding patterns were considered to be the same REA type, or subclone.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Overall genetic diversity. Forty ETs were identified among the 305 L. monocytogenes isolates from patients, seafoods, fish-processing environments, and seawater. ETs identified in our earlier studies (13, 17, 20) were given the same numbers, while new ETs were numbered consecutively after identification. Eighteen of the 40 ETs were represented by a single isolate, and the remaining ETs included the genotypes of from 2 to 71 isolates. The ETs were further divided into a range of REA types. REA types within each ET were designated alphabetically (ET-1a, ET-1b, etc.). The REA nomenclature does not reflect the relatedness of the subclones (a, b, c, etc.) within each ET, as the degree of variation in banding profiles was different among ETs. The banding profiles of the REA types within ET-3 are shown in Fig. 1. No single REA type fell into more than one ET. A total of 103 REA types were distinguished.

View larger version (138K): [in this window] [in a new window]   FIG. 1.   Restriction pattern of L. monocytogenes isolates within ET-3 (a to e) after REA using HaeIII. Lane 1, ET-3a, L 1067 (patient isolate); lane 2, ET-3b, L 1272 (smoked salmon); lane 3, ET-3c, L 1359 (gravad salmon); lane 4, ET-3d, L 3025 (smoked salmon); lane 5, ET-3e, L 3370 (patient); lane 6, ET-3e, L 2127 (smoked salmon). Lanes are numbered from left to right, respectively.

View larger version (32K): [in this window] [in a new window]   FIG. 2.   Genetic relationship among 305 L. monocytogenes isolates from patients, seafoods, seafood-processing environments, and the environment. The dendrogram was based on MEE using 21 enzyme loci. The ETs were further differentiated into REA types (a to k), using the restriction enzyme HaeIII. Abbreviations: N, number of isolates; S, serotype; P, patient; Re, ready-to-eat seafood; R, raw seafood; E, processing environment; E*, processing environment and seawater or seagull; T, transport terminal.

Strains from cases of human listeriosis. The 44 patient isolates belonged to 11 ETs. The most frequent clone was ET-3, which represented the genotype of 12 isolates (28%). Eight patient strains were ET-3e. ET-1 included nine isolates and showed six different REA patterns. Six isolates were ET-11, representing four REA types. ET-6a (seven isolates) was responsible for an outbreak including six patients in 1992, traced to vacuum-packed processed meat (10). The other patient ETs were only sporadically found. Twenty-two of the human isolates were serotype 1, and 22 were serotype 4.

Strains from various Norwegian seafoods and seafood-processing environments. Among the 93 isolates from Norwegian seafoods, seafood-processing environments, and seawater, 27 ETs were identified. The isolates from cold smoked salmon and its processing environment were divided into 16 ETs, the most frequent being ET-33 (six plants), ET-12 (five plants), and ET-3 (four plants). Five subclones were found from two smokehouses, while other subclones were found in only one processing plant.

Isolates from the six processing plants surveyed in detail and from the transport terminal. The distribution of ETs and REA types in the three Norwegian plants where sampling was performed on more than one occasion is shown in Table 1. In SL1, a single subclone, ET-1e, was identified from both sampling days, 9 months apart. On the first occasion, 67% (30 of 45) of samples from the processing environment and 81% of samples from the slaughtered fish were contaminated. Nine months later, after a range of preventive measures had been taken, only 22% (15 of 69) of the environmental samples and none of the 20 slaughtered fish samples contained L. monocytogenes. One isolate collected from seawater outside the processing plant at that time was the same subclone as the processing plant isolates (ET-1e).

Comparison of strains from different sources. Nine ETs were identified both among patient isolates and among isolates from Norwegian products or processing environments. These nine ETs were differentiated into 52 subclones by REA. Nine of these subclones included isolates from patients and seafood products or the processing environment (Table 2).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The 305 L. monocytogenes isolates from patients, seafoods, seafood-processing environments, and seawater characterized by MEE and REA were genetically heterogeneous, and the isolates were differentiated into 40 ETs, several of which could be divided into a range of REA types. The division of the isolates in two clusters of ETs, one (A) including serotype 1 and 4 and NT isolates and the other (B) containing only serotype 1 isolates, is consistent with the results from several other studies of L. monocytogenes isolates of different origins (3, 13, 19).

Four ETs, ET-1, ET-3, ET-6, and ET-11, represented 77% of the isolates from patients. The remainder of the patient ETs comprised only one to three isolates. The most frequently identified clones from patients were also found from different seafoods and processing environments. Six subclones comprised isolates from fish products as well as from patients. The incriminated products were cold smoked salmon, marinated salmon, cold smoked mackerel, and raw fish fillet. Even though no information was available on the food consumption of the patients, the isolation of identical subclones from both sources suggests that contaminated fish products may have been the sources for listeriosis cases in Norway.

Johansson and collaborators (11) identified seven pulsed-field gel electrophoresis (PFGE) types among 22 isolates from retail ready-to-eat fish products from six producers. One PFGE type was found from four producers, while one type was identified from two producers. PFGE type A was identified from seafoods as well as from sporadic cases in Finland. In a Swedish study, two subclones identified by phage typing and PFGE were found from both patients and ready-to-eat fish products (14). Nørrung and Skovgaard (18) identified eight ETs among 20 fish isolates from Denmark, while a total of 33 ETs were found among 245 isolates from different groups of foods and patients. Five of nine ETs from fish were also represented by strains recovered from patients, with 35% of the fish isolates belonging to the clones responsible for more than 75% of the human cases. The isolates, however, were not further subtyped using other methods. Boerlin and collaborators (4) found a range of ETs from imported seafoods, five of nine identified ETs being found from several producers. Some of these ETs had also been frequently isolated from meat and meat products and may therefore be common ETs in Switzerland. No particular ET was recovered from specific fish products. The majority of isolates from fish products in their study belonged to cluster II, equivalent to our cluster B. Boerlin and collaborators concluded that the populations of L. monocytogenes occurring in imported fish and in human patients were different.

Half of the patient isolates in the present study were serotype 1, and the rest were serotype 4. This is consistent with what was reported from Sweden, where human isolates during the 1990s were equally distributed between serotypes 1/2 and 4. In Finland, serotype 1/2 has been recently more frequently isolated from patients (11), and in other countries also serotype 1 seems to have been increasing in the last several years (14).

Some ETs (ET-1, ET-6, and ET-11) were identified from several ready-to-eat foods including smoked salmon, as well as from raw fish. ET-1 and ET-11 were also found from seawater. The frequent findings of ET-1 and ET-6 may reflect the fact that these are the most common ETs in Norway. ET-33 was most frequently identified from smoked salmon and its processing environment (from 6 of 24 smokehouses) and from one slaughterhouse. All ETs from raw fish, filleting environments, and slaughterhouses, except for one, included only one isolate. The diversity of ETs and REA types among isolates from seafoods and processing environments indicates that neither seafoods in general nor smoked salmon in particular offers a niche for specific L. monocytogenes clones.

In several of the processing plants studied in detail, most isolates belonged to a single subclone that appeared to have colonized the plant. In SL1, L. monocytogenes ET-1e was identified from the two sampling dates 9 months apart, showing that this specific subclone had colonized the slaughterhouse. The same subclone was also identified on the second sampling occasion from seawater outside the slaughterhouse, where the transport boats recirculated the water from the fish wells. It is likely that the presence of the ET-1 isolate in the water was due to contamination from the processing plant. Contamination of the processing environment by the seawater is unlikely, because no L. monocytogenes strain was isolated from the living fish coming into the slaughterhouse.

In SL2, subclone ET-36c dominated for almost 1 year, while another clone was identified sporadically for some months. In the Norwegian smokehouse (SmN), the subclone ET-3e predominated and was identified on four occasions in 2 years, while two other clones were found occasionally.

From the non-Norwegian smokehouse SmA, subclone ET-6c dominated on both sampling occasions, 2 months apart. Other ET-6 subclones as well as other ETs were also identified. The contamination pattern was more heterogeneous in the non-Norwegian smokehouse SmB, which was visited only once.

Domination of one clone in the processing plants is consistent with results of a former study where we found that a single ET had colonized a cold-smoked-salmon-processing environment during a period of 8 months and with other ETs being only sporadically identified. Raw fish was shown not to be an important source for L. monocytogenes contamination of the smoked salmon (20). Johansson and coworkers (11) found one PFGE type persisting in a smokehouse for 14 months. Autio and collaborators (1) found that in one rainbow trout smokehouse there were two major contamination sites, the brining and slicing areas, which seemed to be colonized by different pulsotypes. Their results supported our findings that smoked fish is contaminated with L. monocytogenes during processing and that raw fish is not an important source of the bacteria entering the processing environment.

One isolate from a pallet at the transport terminal (ET-54b) was closely related to a subclone isolated from one non-Norwegian smokehouse (ET-54e, SmB), and it was also related to an isolate (ET-54a) from a Norwegian slaughterhouse. This might indicate an earlier common source for the bacteria. The examination at the transport terminal showed that pallets as well as floors of the loading vehicles may be contaminated with L. monocytogenes and thus may be a factor in spreading the bacteria to processing plants. This emphasizes the need for plants to prevent entry of the bacteria into the processing environment when receiving raw material, equipment, etc. Pallets, cases, and other packaging materials should not be taken into the processing premises, and only in-house equipment should be employed for internal storage and transport. In general, preventing the introduction of L. monocytogenes into the processing environment is the most important step in controlling bacteria in the production of seafoods.

In conclusion, the isolation of identical subclones of L. monocytogenes from both human patients and seafoods, including ready-to-eat products, suggests that such products may have been possible sources for listeriosis cases in Norway, although the cases could not be linked to the incriminated foods, as no information on the diet of the patients was available. Ready-to-eat seafoods like smoked fish should, however, be considered risk products for human listeriosis.

Seafoods and seafood-processing environments do not offer a niche for certain L. monocytogenes clones. On the other hand, individual processing plants are likely to be colonized with specific subclones of the bacteria, which may survive in the processing environment for months to years.